______
COMMISSION FOR BASIC SYSTEMS
STEERING GROUP ON RADIO FREQUENCY COORDINATION
GENEVA
16-18 MARCH 2006 / CBS/SG-RFC 2006/Doc. 3.1(4)
______
ITEM 3.1
ENGLISH only
EUMETNET contribution to forthcoming ECC meeting
Interference from RLAN to meteorological radars in the 5 GHz band
(Submitted By P. Tristant, France)
1Summary and Purpose of Document
This document is being presented by EUMETNET to the forthcoming ECC meeting and present evidence of interference from 5 GHz RLAN that occurred to operational meteorological radars in Hungary and Poland.
These elements are compared to similar interference highlighted during testing performed in Canada.
2Action Proposed
For consideration within SG-RFC.
/Doc. ECC (06)XXX
13th meeting
Oulu, 20 – 24 March 2006
Date issued: 6 March 2006
Source: EUMETNET
Subject: Interference from RLAN to meteorological radars in the 5 GHz bandSummary:
This document presents evidence of interference from 5 GHz RLAN that occurred to operational meteorological radars in Hungary and Poland.
These elements are compared to similar interference highlighted during testing performed in Canada.
Proposal:
For consideration within ECC on the justification of the need for mandating Dynamic Frequency Selection (DFS) feature applied to 5 GHz RLAN with appropriate DFS definition in order to ensure protection of Meteorological radars and to possibly find a solution with regards to RLAN based on former ETSI standard.
Background:
1Introduction
Meteorological radars represent key observation stations used for meteorological and environmental measurements and survey, essential to provide short-term alert in case of severe weather conditions (such as flooding, cyclones, hurricanes) endangering populations and strategic economical domains such as transportations, energy or agriculture, as the first line of defence against loss of life and property in flash flood events such as recently in Boscastle (UK) or in several cases in south of France.
There are currently in Europe more than 160 meteorological radars and about 140 in the 5600-5650MHz band for which the detection and monitoring of storms at range relies upon detecting signals just above the environmental noise meaning that even 1dB loss of sensitivity would have a measurable impact on the effective radar coverage.
Following last WRC03 conclusions, related Decision ECC(04)08 and corresponding EC Decision 2005/513/EC, meteorological radars in the 5600-5650 MHz band, as well as all radar types in the 5470-5725 MHz band, have to share the band with RLAN applications under specific regulation such as power limits, power control and Dynamic Frequency Selection (DFS), this latter DFS feature being the main tool that would allow compatibility between RLAN and radars. These RLAN devices are described in ETSI standard EN 301893.
However, it should be noted that the first version of this standard were not specifying adequately DFS feature and it appears now that such equipments are regularly producing interference to operational meteorological radars in these countries, justifying de facto, if needed, the necessity of adequately specified DFS to ensure proper protection of radars in the 5470-5750 MHz band.
Some of these interference events are reported in this documents, also compared to similar interference highlighted during testing performed in Canada.
2Testing performed in Canada
At the last WP 8B meeting (September 05), Canada presented a contribution (8B/293) giving experimental and analytical interference results from 5 GHz RLAN to meteorological radars, concluding on the absolute necessity of DFS.
In particular, experimental studies were performed in a real environment with a RLAN equipment located at 10.6 km from a radar and transmitting a 38 dBm power. It should be noted that, compared with the current European regulation for which a maximum power of 30 dBm is required, this situation is comparable to an RLAN located at roughly 4 km.
Figure 1 (Figure 2 from DOC 8B/293)
Interference effects of 38 dBm RLAN the CWSR98 Radar. RLAN located at Carleton Place, 10.6 Km from the radar and ~ 345 degree Azimuth. There is a streak at ~345 degrees WRT North. The RLAN is located within the streak. Scale circumferential gradations are ~40 Km
The RLAN is located 345° North and this experiment shows interference effect in a number of azimuth, either at the coverage edge or at all distances in the 345° azimuth or in the range 50-90°.
3Interference on operational radars in Hungary and Poland
Meteorological services from Hungary and Poland have been recently alerted by interference events that occurred to their radars that have been shown as being produced by 5 GHz RLAN based on a the initial ETSI standard.
Figure 2
RLAN interference to Legionowo (Poland) meteorological radar (sept 05).
Figure 3
RLAN interference to Budapest (Hungary) meteorological radar (Feb 06).
Figure 4
RLAN interference to Budapest (Hungary) meteorological radar (Feb 06).
These elements shows large similarity with the Canadian testing in that RLAN interference impacts the radar measurements over large azimuth and distances which render radar data totally non exploitable, situation that is clearly critical in zones that are known as being susceptible to flooding.
In the case of figure 2 (Poland), it is worth noting that the impact of RLAN is comparable in rain precipitation levels (dBz) to heavy rain (between 20 and 50 dBz), that will lead to wrong information on the precipitation data used in nowcasting and also in the rain accumulation data that are crucial to survey flooding risks.
The national Polish Radiocommunication administration alerted, it took more than one year to find the source of the interference, mainly due to the weak power of the RLAN signal (radar is much more sensitive than typical measuring tool) It then appeared that this source was a RLAN transmitter operating on frequency with peak very close to the radar frequency and located very close to the radar site.
The RLAN transmitter found by the frequency management authority was equipped with that device but DFS mode was switched off. since the operator has a possibility to switch the DFS device on or off at any time.
It should also be noted that, since the beginning of this year, a similar false echo appeared on the image from Poznan radar site.
Also, in the case of figure 3 and 4 (Hungary), it should be noted that the a number of similar interference events occurred during the last 2 years. The National Frequency Authority stopped these interference last July but the interference recently reappeared in February 2006. In this situation, this radar cannot be used operationally anymore, meaning that precipitation on half of Hungarian territory cannot be monitored.
It should be noted that these images reflects the impact of interference on precipitation measurements but that these interference may have even more higher impact on wind Doppler measurements that are difficult to present or even detect.
4Conclusion
This document presents evidence of interference that occurred to operational meteorological radars in Hungary and Poland from 5 GHz RLAN designed on the basis of the previous ETSI standard.
It is also worth noting that, apparently, similar problems occurred in the Czech Republic as well as in Germany.
It is worth noting that these elements are comparable with similar interference highlighted during testing performed in Canada for RLAN without DFS feature, proving, if needed, the necessity of DFS and the inadequacy of the previous version of ETSI standard in ensuring protection of radars in the 5470-5725 MHz band.
Indeed, the issue is not limited to those countries that have already authorised these old fashioned 5 GHz RLAN equipment since, by principle, these RLANs benefit from a free circulation within the EU and, even though not authorised in used in some countries, could be placed on the market in these countries.
Meteorological radars take part of processes related to safety of life and safeguarding strategic European economical domains that results in short-term alert in case of severe meteorological and environmental events such as flooding, severe storms or even pollution monitoring. These radars cannot be put at risk with an uncontrolled deployment of unlicensed radio equipment as RLAN, at least if these equipment are not compliant with the up-to-date EC and ECC regulations and ETSI standard.
On this basis, these elements are presented to ECC in a view to justify the need for mandatory Dynamic Frequency Selection (DFS) feature applied to 5 GHz RLAN, under the specification described in the last ETSI standard version (EN 301893) and to possibly find a solution with regards to RLAN based on former ETSI standard.
------